The goal of the proposed research is to understand the capacity of persons with post-stroke hemiparesis to adapt locomotor interlimb coordination and the influence of this on gait symmetry. Utilizing this knowledge, the applicant's long-term research goal is to develop scientifically-based therapies to advance the physical rehabilitation and recovery of locomotion in persons following a stroke. A five-year research and training plan has been developed to meet these goals. This plan consists of training in the neural mechanisms of locomotion, motor adaptations in persons with neurological damage, gait measurement and analysis and the design and implementation of randomized clinical trials in rehabilitation. This plan will be implemented under the mentorship of the primary mentor, an expert in motor adaptation in persons with neurological damage and the cosponsor, an expert in the analysis of walking in persons with gait dysfunction and in the design and implementation of clinical trials in rehabilitation. Training will consist of structured readings, attendance at course lectures and seminars, structured laboratory experiences and presentation at academic meetings. With complete support from the candidate's department chair, the candidate will have access to two fully equipped laboratories capable of measuring 3-dimensional human movement and force production. The proposed research will test the general hypothesis that persons with post-stroke hemiparesis are able to adapt their locomotor patterns following walking on a split-belt treadmill, where each leg is moved at a different speed. It is hypothesized that the adapted pattern of interlimb coordination following split-belt treadmill locomotion may be used to improve gait symmetry and function in persons post-stroke. Specific aims will test: 1) the capacity of persons with post-stroke hemiparesis to adapt interlimb coordination, 2) the generalization of adapted patterns to overground locomotion and, 3) the feasibility of using split-belt treadmill training to improve gait symmetry in persons post-stroke. Relevance: Of the 700,000 persons affected by stroke each year, 50% will continue to demonstrate long term walking dysfunction, some caused by poor coordination between the legs. This research investigates the potential of a novel type of treadmill walking to influence coordination between the legs in persons who have sustained a stroke.